Bachelor thesis Chemical Engineering

(1)

Bachelor thesis Chemical Engineering

Coal as aid for the removal of hormones from the effluent of a WWTP and an insight in the micro-plastics problem

Author: J.A. Kabel Eeldersingel 28 9726 AR Groningen S2036908

14-06-2013

Revised: 05-07-2013

Under supervision of: M.W.M. Boesten and F. Picchioni

(2)

2 Preface

This research is the final part of the bachelor program of Chemical Engineering at the Rijksuniversiteit Groningen (RUG). The project started 22 april 2013 and took 3 months to complete. The given

assignment has been a reflection of the skills and knowledge acquired in the bachelor program.

I would like to thank M.W.M. Boesten for his guidance and criticism, A. Haijer from Water and Energy Solutions for extra assistance, A.C. Meinema for giving access to hormones and last but not least I would like to thank J.H. Marsman for help with the analysis and the experimental overall, Jasper Kabel

Abbreviations

WWTP – Waste Water Treatment Plant PE – Polyethylene

PP – Polypropylene

STOWA – Stichting Toegepast Onderzoek Waterbeheer LOES – Landelijk onderzoek Oestrogene stoffen

DSC – Dry Solid Content OBD – Overall Block Diagram PFD – Process Flow Diagram Commonly used terms

Influent - Wastewater flowing into a treatment plant Effluent -Treated wastewater

Estrogen – female hormone

(3)

3

1. Introduction

1.1 Introduction WWTP Garmerwolde

The Waste Water Treatment Plant (WWTP) situated in Garmerwolde is part of the ‘Waterschap Noorderzijlvest’. The plant cleans the sewage water from Groningen and surroundings with mechanical, biological and chemical treatments. The result is clean water, which is released in the Eemskanaal, and sludge. The sludge is processed in the WWTP to remove as much water and

hazardous chemicals as possible. The processed sludge is then transported to Swiss Combi for further drying. Swiss Combi is separate company located on site. They sell sludge granulates to ENCI in Maastricht. ENCI burns these granulates to produce energy.

The Rijksuniversiteit Groningen (RUG) has been collaborating with ‘Waterschap Noorderzijlvest’ for a few years now. This collaboration provided the Waterschap with possible solutions and

improvements, while students had the possibility to their thesis on a subject related to the Waterschap. The first students were Gijsbert Haaksman (2009) and Martin meelker and Olivier Burgering (2010), who did their master thesis about various subjects related to the WWTP

Garmerwolde. These thesis’s were quite successful, resulting in more opportunities for students in both the master and bachelor trajectory. This year (2013) Marc Meijerink, Marthe Sveistrup, Henrieke Heideman, Machiel van Essen and I are doing our bachelor thesis in collaboration with Garmerwolde. Although my assignment is not directly related to the WWTP in Garmerwolde it can give insights for the future.

1.2 Objective research and scope

The objective of this research is to look at the possibility of coal to decrease the concentration of certain hormones and to look in to the micro-plastic (micro PE, micro PP etcetera) problem.

The choice to use coal comes from earlier research which shows that coal improves the dewatering and can be used for flocculation (Meelker, 2010) and because the EDCAT project in Great-Britain showed that active coal improved the effluent on hormonal level (STOWA Lahr & de Lange, 2009), thus coal could be interesting as well. It will not be a research based on the circumstances at a WWTP, because it is not sure if coal can decrease the concentration of hormones. This means other parameters need to be ruled out. The coal used is not activated and thus if adsorption is proved with the experimental, the adsorption will be expected not to be that great. Even with low adsorption rates, the coal will still be viable thanks to the large amounts of coal that are available as well as the relative low costs of it.

The micro-plastic problem is looked at from the Garmerwolde WWTP with the assumption that the micro beads go through the system fairly unhindered. The problem will be looked at purely

theoretical and the possibility to combine the solution to the micro PE problem and the concentration decrease of hormones will be sought.

1.3 Assignment

To address this objective, an assignment was formulated:

At laboratorial scale the effectiveness of coal as filter aid for the removal of the hormone Bèta- estradiol is tested and the micro-plastic problem is addressed by looking at the Garmerwolde WWTP.

After these two goals the possibility of combining the solution to the micro-plastic problem and the coal as filter aid will be sought. A literature study is done on the micro-plastic problem, the hormone problem and on coal adsorption in general.

The practical work should be reproducible.

(6)

6 1.4 Problem definition

In the literature study part the problems given by the hormones, medicine and micro-plastics will addressed.

The definition of the removal of Bèta-estradiol will be:

In this case removal means that the material exists in the influent and will exist less in the effluent.

This means the material is not removed from the environment, but just from the waste stream. The removal will be a result of either filtration or adsorption. Because the WWTP sludge is burned in the Netherlands, the hormones (in this case just Bèta-estradiol) will not be available for the environment.

Many parties are involved, they are called stakeholders.

Stakeholders

Society: Both one of the reasons of the hormone, medicine and micro-plastic pollution and a victim.

The victim role can be seen in several ways. First of all the hormones and medicine could be harmful for people, secondly a lot of people are not aware of what kind of pollution certain products give.

This asks for campaigns to raise awareness, so people know what kind of pollution they bring in the environment themselves.

WWTP’s: The Waste Water Treatment Plants should be the ones cleaning the water. Important is that innovation should be sought all the time to keep the risks of pollution coming in the surface water as low as possible. The WWTP’s should be involved with the awareness campaigns, because the WWTP’s measure the concentrations of their outlets, knowing what still pollutes the

environment.

Green organizations: For example Noordzee and the Plastic Soup Foundations. These organizations try to raise awareness for the environment, in some cases with success. The Plastic Soup Foundation addressed several cosmetic companies about micro-plastics, resulting in some positive responds by banning micro-plastics from products by Colgate, Palmolive, Loreal and Beiersdorf (Noordzee, 2013).

Health care industry: The companies that produce medicine, hormones and use micro-plastics. These companies primary goal is to make profit, which can result in violation of legislation about the concentrations medicine and hormone in waste water. These companies most likely clash with the green organizations. But they do care about their image (bad image means less profit most of the time), resulting in some cooperation with the green organizations and government.

Government: The government uses legislation to keep pollutions at decent levels. The legislation is designed in cooperation with researches and science institutes to adjust the parameters and decent levels for both the industry and the environment. The government also uses campaigns to raise awareness so that society will realize what harm certain products can do.

Institutes: Institutes like STOWA research and measure valuable parameters to enhance the

knowledge about waste water and their treatment plants. The institutes mostly provide knowledge and lack real power to enforce changes at the WWTP’s and companies.

(7)

7 1.5 Research schedule

Literature study

Before the practical work, a literature study will be done. The study will focus on:

Coal

The overall coal filtration and adsorption process will be studied, along with the differences between normal coal and active coal. The information will be gathered from scientific researches and STOWA reports.

Micro PE

The problem the micro PE gives will be looked, along with a look at the reasons why micro PE was implemented in products in the first place. The sources will be STOWA reports and reports from Green Organizations.

Hormones/medicines

The problems and uncertainties the hormones and medicines give will be studied. The basis of the research will be reports from STOWA, which provide information from the branch.

Experimental

The experimental work is focused on decreasing the concentration of a hormone in a solution with the help of coal.

The experimental work will not be a small scale replica from the actual WWTP. No sludge and such will be used, because the focus lays at the study if lowering of the concentration of the hormone is possible, no other factors should have a chance to play a role.

The influent will be a solution of a hormone.

First the smallest particle size coal is used, because it is expected to have the best adsorption possibilities. If interaction has been proven, a random sample of the coal will be used with an

expected lower decrease of the concentration of the hormone. Also the time and the amount of coal will be varied, with for both parameters an expected larger decrease of the concentration if their amount is increased.

Taking a sample of the smallest particle size will affect the PSD of the batch coal used. Therefor a new batch of the coal will be taken for the experimental with a random sample of coal.

(8)

8

2. Theory

2.1 Hormones

The last 15 years there have been more attention to the influences of hormones on the environment.

Hormones can disrupt an organism in several ways (STOWA Lahr & de Lange, 2009):

● The hormones induces an error in the genitalia, also known as intersex

● Changes in the sex distribution of an organism

● Unwanted production of the female protein vitellogenin in the male

These three disruption ways will in principle lead to a disrupted reproduction, which will result in a lower population growth.

The hormones can disrupt the organism via four different effects, shown in table 1.

Effect Mechanism Result

Estrogen effect Stimulated estrogen receptor Feminization Anti-estrogen effect Blocked estrogen receptor Defeminization Androgenic effect Stimulated androgenic receptor Masculinization Anti-androgenic effect Blocked androgenic receptor Demasculinization

Table 1 Hormone disruptions

The most attention in studies goes to the female hormone group of estrogens. The estrogens can be divided in four groups (STOWA Lahr & de Lange, 2009):

●Natural estrogen hormones: These are produced by vertebrate animals

●Synthetic estrogen hormones: Hormones produced for anti-conception or to enhance the feminine fertility

●Xeno-estrogen hormones: Foreign chemical substance that exhibit a certain degree of estrogen behavior

●Fyto-estrogen hormones: Substances from botanical origin that exhibit estrogen behavior A big study in the Netherlands called ‘Landelijk onderzoek Oestrogene stoffen’ (LOES) proved that certain hormones exist in surface waters. The results from LOES are shown in appendix A.

So it is established that hormones do exist in waters in concentration in the order of ng/l to μg/l. The question now is, are these hormones harmful and if so in what concentrations.

The first suggestions that a change in hormones could be harmful comes from a Danish study from the early ’90. The study showed that the sperm quality of men was dropping. This study was followed up by the book ‘our stolen future’ (colbern, 1996) in which alarm was raised because of the drop of fertility of both men and animals due to hormone disrupting substances.

A nuance to the alarm came from the ‘Gezondheidsraad’ (a Dutch instance), which concluded that exposure to hormones/hormone disrupting substances are not a direct threat for the public health.

The different opinions are logic, because the direct effect of exposure to a hormone or hormone disrupting substance is hard to determine. Overall there is no hard proof that the exposure to hormones is indeed harmful for humans, for example even the highest measured concentration of ethinyl estradiol is approximately 3500 times below the normal doses in the anti-conception pil (STOWA Derksen & Lahr, 2003).

This does not mean we can dump the hormones and hormone disrupting substances anywhere we want, because damage to several organisms has been proven (STOWA Hekster & Mons, 2004) (STOWA Lahr & de Lange, 2009).

(9)

9 Snails

Lab research to several fresh water snails proved feminization due to hormones including ethinyl estradiol and Bisphenol A. The concentrations that were needed to see an effect were in the ng/l range. The concentrations found with LOES were in the same range.

No field research has been done (yet), because the research is labor-intensive because the snails must be picked out of their shell.

mussels

For mussels the fresh water mussel (Elliptio complanata) has been studied well in Montreal, Canada.

The fresh water mussel was studied at a WWTP were they compared the percentage female mussels before and after the WWTP. The difference between before and after was found to be quite

significant. Before the WWTP 41% of the mussels was female, but after the WWTP 66% was female.

Rainbow trout

In the nineties a study in Britain with rainbow trout was done. The trouts were put in cages and these cages were let in mobile flow-through systems. In these systems the trouts were exposed to different dilutions of effluents. After three weeks the trouts were examined. A heighted concentration of vitellogenine (VTG) was found in the blood of the males. VTG is a yolk protein that is normally produced in the liver of the females. VTG will accumulate in the males, because in contrary to the females the males do not have a natural way of losing it. The females lose the hormone when they lay eggs.

Another interesting thing found, was that female ovules where found in the tissue of the male genitalia.

Fathead minnow

A study in Canada for three years showed the effect of 17-alfa-ethinylestradiol (one of the

components in the anti-conception pil) on the fathead minnow (Pimephales promelas). This fish was almost extinct at the test area after maintaining a constant concentration of 5-6 ng/l. Before becoming extinct the males had the same symptoms as the rainbow trout (VTG and female ovules).

The females had impaired genitals and a slowed egg production. This study was unique in a sense that it was the first study showing that estrogens had effect on a population level of organisms.

A side note worth mentioning is that a few years after stopping the exposure to 17-alfa- ethinylestradiol the survivors of the fathead minnow population were able to create a new population.

Conclusion

Although there are no studies that show direct harm to humans, harm to several organisms has been proven. Harming organisms will disrupt the balance in nature, which can backfire to the humans in a later stadium. And for a future perspective, if hormones are kept being let in the surface waters the concentrations will increase which could give a harmful situation for humans in the future.

(10)

10 2.2 Medicines

Hormones are not the only materials that might cause problems for organisms. A broad spectrum of medicine has been found in the surface waters as well. Most medicine have to properties that they are soluble in water, mobile and that they work at low concentrations. These three properties give problems in the environment, because the medicine problem will spread out quickly and will still give problems due to fact that the hormones will work at low concentrations.

The first time attention was paid to medicine in the environment was in the early eighties. A few studies (de Roij & de Vries, 1982; van der Heide & Hueck-van der Plas, 1982; Watts, 1983: Richardson

& Bowron, 1985) looked in to the matter. But it was not till 1992 when it the attention became serious. In Germany an unknown substance was found while screening ground water for herbicides and pesticides. The unknown substance was a breakdown product from fibrates (see table 2). Further research showed that these breakdown products were provable in a lot of waters. These results resulted in new studies in Denmark, Germany, Switzerland and the Netherlands measuring concentrations in influents, effluents, several rivers, the North sea and drinking water (STOWA Derksen & Lahr, 2003) (STOWA Derksen & ter Laak, 2013).

One of the organizations in the Netherlands is the RIZA/RIWA.

According to the RIZA/RIWA approximately 12000 human medicine and 2500 animal medicine are allowed. For the environment the active medicine are the most interesting, resulting in a shortlist of 850 humane medicine and 250 animal medicine. From the humane medicine around the 200 are monitored. Table 2 shows the important groups of humane medicines, table 3 show the important animal medicine (STOWA Hekster & Mons, 2004) (KWR STOWA Derksen & ter Laak, 2006).

Group Medicine Application Example of active compound

Analgetica Painkillers Aspirin, ibuprofen

Antibiotica Treatment of infections Amoxicillin, erythromycin

Anti-epileptica Treatment of epilepsy Carbamazepine

Béta-blocker Treatment high blood pressure and heartproblems

Metoprolol, sotalol

Cytostatica Treatment of cancer Cyclofosfamide

Fibrates Fatregulating Bezafibrate, fenofibrate

Psychofarmaca Treatment of depression Diazepam

X-ray contrast material X-ray diagnostics Amidotrizoineacid, iopamidol

Table 2 human medicine

(11)

11 Group Medicine Compound groups Example of active compound

Antibiotics Aminoglycosiden

Bèta lactam derivatives Diaminopyrimides Macroliden Quinolen Sulfonamiden Tetracyclinen

Neomycine, Spectinomycine Amoxycilline

Trimethoprim

Erythromycine, Tylosine, Enrofloxacin, Flumequin Sulfachloorpyridazine Doxycycline, Oxytetracycline

Antiparasitica - Flubendazol, Ivermectine

Coccidiostatica - Monensin, Nicarbazine,

Salinomycine

Table 3 Animal medicine

On the next page (figure 1) a few monitored medicine are shown. In the figure the removal percentage from the waste water stream is given. The percentage is an average of eight WWTP’s.

The n value is on how many measurements the average is based. The standard deviation is shown as well. For some medicine the removal percentage is quite high, but not one medicine is removed completely, this means the stream leaving the WWTP will always contain medicine.

In appendix A. the results of LOES are three tables with maximum measured concentration of several medicine. Both figure 1 and these results show that medicine will come in the environment. Same as with the hormones, the question is do the medicine harm and if so in what concentration.

The knowledge about the effects is increasing, but a lot is still unknown. A problem with the studies is that the concentrations in the surface waters are a lot lower than the concentrations used in the lab. Another problem is that in the lab mostly the short term effects are measured, while in the environment organisms are exposed to the medicine for a long time to a mixture of a lot of medicine.

To judge the risks it is important to look at the long term effects and the combination of several medicine. A third problem is that the effects of the medicine are not always measurable with standard biological measure methods.

For humans the expectation is that the low concentrations of medicine will not have a harmful effect on humans, mainly because the concentrations are around a factor 10⁶ lower than the doses

prescribed. The only small effect some medicine can have is an allergic reaction, because really small amounts can evoke a reaction. Most of the time these reactions disappear when exposure to the medicine is stopped. Most of the medicines that evoke an allergic reaction are antibiotics.

These antibiotics have another problem. They are used against bacteria’s, but bacteria’s have to tendency to become resistant if exposed to the antibiotics, because the bacteria’s with the highest resistance to the antibiotics will survive. If this happens in nature all the time, none of the antibiotics will be working well. This can give major health issues on global scale in the future (STOWA Derksen

& Lahr, 2003) (KWR STOWA Derksen & ter Laak, 2006).

For other organisms the medicine can give problems at this stage and the concentrations that are now measured. The medicine are labeled with an EC50 concentration. The EC50 concentration is the half maximal effective concentration, which is the concentration of a drug, antibody or toxicant which produces 50% of the maximal possible response for that drug, antibody or toxicant. The lower this EC50 concentration, the more toxic the compound is. In graph the results of a study from Derksen (2001) are shown. The study was a collection of Eco toxicological data of about 120 of the most used humane medicines. The result is a distribution of the number of medicine per EC50 concentration, which is shown in graph 1. When the EC50 concentration is larger than 100 mg/l the compound is considered not toxic, between 10 – 100 mg/l the compound is considered harmful and an EC50

concentration smaller than 10 mg/l means a compound is toxic. As can be seen in graph 1 several

(12)

12 medicine can be considered at least harmful. But because the concentrations measured are in the order of ng/l or µg/l no direct harm is expected, although chronic negative effects have to be considered as possible outcome (STOWA Derksen & Lahr, 2003).

And on a longer term the medicine can accumulate in different kinds of tissue of organisms. This does not give direct problems for these organisms themselves, but when they are consumed they can poison their predator or when they die they can release a relative high dose of medicine in a local area. The precise damage this causes is unknown and it is hard to be studied precisely.

For some organisms some specific effects are studied and proven (STOWA Derksen & Lahr, 2003) (STOWA Hekster & Mons, 2004):

●Extra sensitivity for algae and cyanobacteria for antibiotics

●Extra sensitivity for algae for beta blockers

●Kidney damage caused by diclofenac for humans

●Mass come out of ovums and spermatozoons by mussels caused by fluoxetine

●Disruption of the skinning of lobster-like animals

Graph 1

Conclusion

It is expected that humans will not have direct negative effects from the medicines found in surface waters at the concentrations at which they are measured. However some effects for other organisms are suggested and proven which means action should be taken to at least prevent an increasing concentration of the medicines. In a future perspective, preventing an increase of the concentrations is essential to prevent the concentrations from getting in the concentrations ranges in which they might have effect on humans.

(13)

13

Graph 2 Removal percentages of medicine

(14)

14 2.3 Coal

In principle coal as filter aid will work via adsorption. Adsorption is a heterogeneous reaction where a particle in gas or liquid phase(the adsorbate) is bound to a solid or liquid called the adsorbent. In this way the particle is removed from the gas or liquid phase. This process creates a layer of the adsorbate on the adsorbent. The limit of adsorption lays in the adsorbent, because of this layer forming. When in theory all the surface of the adsorbent is covert, no more adsorption can take place. After adsorbing, the adsorbent has to be destroyed or regenerated. Active coal is the most used form for coal filtration (Glossary, 2009) ((wordnet), 2011) (Anon., 2009).

Adsorption is described via isotherms. Isotherms are the amount of adsorbate on the adsorbent as a function of its pressure (if gas) or concentration (if liquid). By normalizing with the mass of the adsorbent comparison between different materials is possible.

There are different ways to describe the isotherms:

●Linear adsorption isotherm

●The Freundlich adsorption equation

●Langmuir isotherm

●BET Theory

●Kisliuk model

Linear adsorption isotherm

The simplest way to describe the isotherm is the linear adsorption isotherm or Henry adsorption constant.

For gasses:

For liquids:

Where: X = The fraction of surface covered by the adsorbate (-) KH = Henry’s adsorption constant (1/atm or l/mol) P = Partial pressure adsorbate (atm)

C = Concentration adsorbate (mol/l)

The linear adsorption isotherm can be used to describe the initial part of many practical isotherms. It is typically taken as valid for low surface coverages, and the adsorption energy being independent of the coverage (Yildirim, 2006).

The Freundlich adsorption equation

This equation is a curve relating concentration of a solved material on the surface of an adsorvent to the concentration of the solved material in the liquid.

For gasses:

Or For liquids:

Or

(15)

15 Where: X = Mass of adsorbate (gr)

m = Mass of adsorbent (gr)

p = Equilibrium pressure of adsorbate (atm)

c = Equilibrium concentration of adsorbate in solution (mol/l)

K and n = Constants for a given adsorbate and adsorbent at a certain temperature (1/atm or l/mol)

The adsorption will become independent of the pressure at high temperature, because for high pressures.

The Freundlich adsorption equation is used when the actual identity of the solved material is unknown (Anon., 2008).

Langmuir isotherm

Irving Langmuir created this model that relates the coverage or adsorption of molecules on a solid surface to gas pressure or concentration above the solid surface at a fixed temperature. It is the most common used isotherm equation thanks to its simplicity and its ability to fit a variety of adsorption data. This model is based on four assumptions:

●All of the adsorption sites are the same and can only fit one molecule

●The surface is energetically homogeneous and there is no interaction between adsorbed molecules

●The are no phase transitions

●At the maximum adsorption point, only a monolayer is formed. The adsorption only occurs on localized sites on the surface, not on other adsorbates.

When these four assumptions are met, there is an ideally situation. Unfortunately there are always imperfections on the surface, adsorbed molecules are not per se inert and the adsorption mechanics will differ for the first molecule compared to the last. The fourth assumption gives the most trouble, as always more molecules will be adsorbed than just the amount to form a monolayer.

This isotherm is most used in applications in surface kinetics and thermodynamics.

The model suggests that adsorption takes place via the following mechanism:

Where: A = A molecule S = An adsorption site

The kinetic rate constants are k and k^-1. The surface coverage as the fraction of the adsorption sites occupied at equilibrium will be defined as θ. This gives:

Or

Where: P = The partial pressure of the gas or the molar concentration of the solution (atm) This model gives at low pressures , and at high pressures .

θ is hard to measure experimentally, thus it is replaced by . This is the quantity in moles or grams adsorbed at standard temperature and pressure per gram of adsorbent (v) divided by the volume of material that’s getting adsorbed to form a monolayer on the adsorbent at standard temperature and pressure (vmon).

(16)

16 This gives the expression:

(Anon., sd) BET Theory

The BET theory is a modification of the Langmuir isotherm taking multilayer adsorption into account.

The mechanism becomes:

Etcetera.

The derivation of the formula is more difficult than with the Langmuir isotherm, but for gasses one finally obtains:

Where: x = Pressure divided by the vapor pressure for the adsorbate (-)

v = volume of adsorbed adsorbate at standard temperature and pressure (cm³)

vmon = amount of adsorbate required to form a monolayer at standard temperature and pressure (cm³)

c = The equilibrium constant K used in Langmuir isotherm multiplied by the vapor pressure of the adsorbate (-)

(Brunauer, et al., 1938) Kisliuk model

This model is an alteration of the Langmuir isotherm. The basis from the alteration comes from the observation that adsorption is more likely to occur around already adsorbed molecules on the surface of the adsorbent. This renders Langmuir’s isotherm ineffective for the purpose of modeling.

Paul Kisliuk thought of a precursor state theory to compensate for the increasing probability of adsorption around molecules present on the adsorbent surface. This theory states that molecules enter a precursor state at the interface between the solid adsorbent and the adsorbate. From this state the adsorbate either adsorbs to the adsorbent or desorb back in the gas or liquid.

Implementing the precursor state theory the following formula is obtained:

Where: θ = The fractional coverage of the adsorbent with adsorbate (-)

R’ = Rate constant, represents the impact of diffusion on the monolayer formation and is proportional to the square rot of the system’s diffusion coefficient (s^-1)

t = The immersion time (s)

kE = This is the sticking coefficient, which is , with SD is the adsorption rate constant, kES is the rate at which the adsorbate desorbs and SE is the desorption rate constant (-)

(Sivaraman, et al., 2009)

(17)

17 For this bachelor thesis not active coal, but normal coal is used. Normal coal is likely to be significant less active as an adsorbent. This comes because of the one big difference between ‘normal’ coal and active coal. The big difference lies in the structure, the structure of active coal creates an internal surface in the order of 500 – 2000 m² /gram coal. The internal surface of normal coal is in the order of 50 m² / gram coal. (Thomas & Damberger, 1976)

Commercially there are two version of active coal, powder and granulate version.

(Dąbrowski, et al., 2005)

2.4 Micro-Plastics 2.4.1 Problem

Micro-Plastics are microbeads of plastics (for example polyethylene or PE, polypropylene or PP and polyethylene terephthalate or PET) which are smaller than 5 mm (see image 1 for an example).

Microbeads are uniform polymer particles which are used in a wide area of applications. The micro- plastics are popular because of their polishing and emulsifying ability and the ease at which the properties can be adjusted to fulfill the desirable function. The three properties that are adjusted the most are density, color and roughness.

Image 1 Microbeads

For several decades the micro-plastics are used, but the possible problems where first addressed in 2004 by Richard Thompson. In his article in Science he was wrote about the presence of micro- plastics on beaches and in water columns of the North Sea. In 2011 the first recap of the scientific work addressing the micro-plastics was made by Leslie. The conclusion was that still little is known about the amount of micro-plastics used and how much micro-plastics end up in the surface waters via WWTP’s. The first estimate of how much micro-plastics people use came not long after the recap by Gouin (2011). This estimate was about polyethylene alone and was only about polyethylene from liquid soaps. Gouin came at an average of 2.4 mg per person per day.

(Roex, et al., sd)

In 2012 a study in the Netherlands measured the concentration of micro-plastics at the WWTP in Heenvliet. The influent had an average of 200 micro-plastic particles per liter, the effluent contained round 20 particles per liter. This means 90% of the micro-plastic particles do get filtered out.

This seems a rather small number, but with estimated 2 milliard cubic meters purified every year in the Netherlands, the amount of micro-plastics that end up in the surface waters is:

particles / year

(18)

18

Or: particles / day

The question is what harms can and will these plastics do. These plastics are stable, because it can be assumed that the most unstable variants are part of the 90% of the micro-plastics that get filtered out. This means the plastics will remain in the environment for at least decades, because the

degradability lies in that order or even higher. This means there will be an accumulation of plastics in the environment in the years that are coming. This still would not have been a big problem if the plastic were completely harmful, but the opposite is true. Several studies show different types of ways the micro-plastics can exert certain damage or be a threat for organisms (Roex, et al., sd).

For instance a literature study by Deltares and IVM shows that humans and animals are able to absorb micro-plastics in their tissue and bodily fluid. After being absorbed, the particles can be taken in the digestive system. Via the digestive system the particles are able to reach the cardiovascular system and the lymphatic system. There the micro-plastics could cause local infections and changes in gene expression. Wick (2010) proved that polystyrene particles with a maximum size of 240 nm can be carried from mother to child via the placenta (Roex, et al., sd) (Appie, 2012).

This in combination with the ability of the micro-plastics to absorb certain other materials gives a broad spectrum of problems the plastics can give. The plastics can have these materials absorbed by choice of design or by absorbing them later. This means the plastics can introduce chemical

contaminations and pathogens in organisms that were in initially in the plastic by design, but it is also possible that chemical contaminations that exist in for example surface waters like pesticides are absorbed and then introduced in an organism (Appie, 2012).

An example of a specific effect on an organism comes from a study from the United States by Bhattacharya in 2011. This study proved that the micro-plastics have a negative effect on the photosynthetic capability of green algae (Leslie, et al., 2012).

A system that can be used to take small particles out of a solution is a DAF. DAF stands for Dissolved Air Flotation. The removal is done by dissolving air in the water under pressure. This air is released at atmospheric pressure in a flotation tank or basin. The air will form tiny bubbles which will adhere to the small particles. The small particles will float to the surface where the can be removed. A baffle will prevent the floating material to continue in the system (Kiuru & Vahala, 2000).

Image 2 DAF

(19)

19 The DAF is not used for micro-plastics mainly because the concentration of the micro-plastics is fairly low. Considering the DAF will be installed at the end of the WWTP system the concentration will be around 20 particles per liter. The DAF will be very inefficient with such low amount of particles (Kiuru

& Vahala, 2000).

Conclusion

The attention that micro-plastics and the possible problem is getting is significantly increased the last decade. The first reports about the effects and possible effects of the micro-plastics justify this.

Although a lot of studies are still running, the few studies that show result all point in the same direction.

Because the micro-plastics are stable, the only chance to keep the thread away is by taking the micro-plastics out of the water. This brings the WWTP’s in an important role. Although, if most WWTP’s do not differ to much in effectiveness, the WWTP’s have an effectiveness of ~90%,

everything that can improve this number should be considered. A special deeper look can be done on DAF systems, although they do not seem viable for the relatively low concentration of micro-plastics the DAF still could be interesting if it is combined with the removal of other problematic particles as well (Leslie, et al., 2012).

2.4.2 Micro-plastics and Garmerwolde

The WWTP Garmerwolde has to filter as much micro-plastics out as possible, like any other WWTP.

The area Garmerwolde manages is 144000 hectares big, this covers approximately 375000 people.

If the earlier mentioned average of Gouin (2.4 mg per person per day (Roex, et al., sd)) is used, a total mass of gr micro-PE has to be removed from the influent every day.

Of course this is only micro-PE and only from liquid soaps. If considered that the three most used micro-plastics (PE, PP and PET) are the majority of the micro-plastics and if PP and PET are used at a same amount as PE, the total mass micro-plastics per day will be 2700 gr. This is still only from liquid soaps. Other sources are toothpaste, skin care products, paint etc. The amount of micro-plastics these other sources produce will be considered roughly the same as for liquid soap. This gives a total mass of 5400 gr. micro-plastics per day that have to be removed from the influent.

The total amount of water that is processed at Garmerwolde is 29000000 cubic meters per year (STOWA, sd). This means 79000 cubic meters per day. This gives a concentration of micro-plastics of

µg per liter in the influent. Considering the ~90% removal effectiveness, the effluent will have a micro-plastic concentration of 6.8 µg per liter. The removed micro-plastics will be in the slib, which will be burned in a later stage. This means this plastic has no chance to come back in the environment.

(20)

20 2.4.2.1 Overall schematic

Image 3 is the OBD (overall block diagram) of the WWTP focusing on the hormones and micro PE.

WWTP Garmerwolde

Sludge cake -5736 kg/h -25.87% DSC -10 °C -10.01325 bar -Sludge (S) -Contains:

-Micro-plastics 0.18 kg/

h

-Hormones 0.048-0.41 g/h

Extern sludge -22831 kg/h -3.2% DSC -10 °C -1.01325 bar -sludge(L)

H&Aa sludge -13575 kg/h -3.7% DSC -10 °C -1.01325 bar -sludge(L)

Influent -2916667 kg/h -12.9 °C -1.01325 bar -Water (L) -7.5 pH -Contains:

-Micro-plastics 0.20 kg/h -Hormones 0.050-0.44 g/h

FeCl3 -130 kg/h

-40 wt% FeCl3 in water -10 °C

-1.01325 bar

Poly-electrolyte -15 kg/h

-1 wt% Poly-electrolyte in water

-10 °C -1.01325 bar

Biogas -251 m3/h -30 °C -1.01325 bar -Gas (G)

Effluent -2947481 kg/h -12.9 °C -1.01325 bar -Water (L) -7.1 pH Contains:

-Micro-plastics 0.020 kg/

h

-Hormones 0.002-0.029 g/h

Image 3 overall schematic of Garmerwolde

The percentage of the micro-plastics is based on the calculations earlier. It is assumed that the smallest micro-plastics will leave the WWTP with the water, because the smaller particles will be more likely to float instead of sediment.

The percentage of hormone is based on the concentrations from β-estradiol from the LOES (STOWA Derksen & Lahr, 2003). The raw waste water concentrations vary from 17-150 ng/l, while in the effluent the concentration is between 0.8-10 ng/l.

(21)

21 2.4.2.2 Detaillistic schematic

Image 3 shows a PFD (process flow diagram) of the Garmerwolde WWTP. As earlier assumed the smallest/lightest micro-plastics will float. A ton is equivalent to 1000 kg.

Image 4 Detailistic schematic of Garmerwolde (Stoffelsma, 2012)

●The influent has 68 µg per liter micro-plastics, this is considered 100%. Following the trail of the influent the micro-plastics will first pass the mechanical filter. Only the biggest micro-plastics might be filtered out here. Assumed is that 60% of the micro-plastics will continue, this leaves a

concentration of 40.8 µg per liter in the stream.

●The sand remover will not remove micro-plastics. The sand remover is based on density and the density of the micro-plastics is significantly lower, meaning that the micro-plastics will not sediment.

This means the concentration of 40.8 µg per liter still remains in the stream.

●The reflux stream will increase the micro-plastic concentration, because of the way it is added. The adding can be compared with pouring, which means the floating particles will be poured out. The amount added is the same as will be mentioned at the reflux part later.

●Next the plastics will pass the two sedimentations tanks. Only the heavier micro-plastics might sediment here, but most of the bigger/heavier micro-plastics have already been removed from the stream. Still there is assumed roughly 20% of the initial concentration will be lost here because some micro-plastics will be captured in the sludge.

(22)

22

●Next the reflux will decrease the concentration by 75%. This value will be added to the stream earlier (see point 3). This drops the concentration to about 7.3 µg per liter in the initial situation. This stream leaves the WWTP, meaning the effluent also has a concentration of about 7.3 µg per liter.

This process has a big flaw on the micro-plastics area. In the initial situation the outcome will be 10%

of the ingoing concentration, but because of the reflux stream there will be an accumulation of micro-plastics. This comes because the reflux stream takes 75% of the concentration of the stream just before the reflux. Even if the ingoing concentration of the system remains the same, the

concentration of the stream just before the reflux will keep increasing. This means the absolute value of micro-plastics in the effluent will increase as well and thus the system needs adjustments.

(23)

23

3. Experimental 3.1 Setup

The first problem was how to test the adsorption of hormones. After discussing with my supervisor Boesten the choice became a delta measurement. This means that not the direct adsorption to coal is measured, but the differences in the concentration of hormones before and after exposure to the coal.

This gave the general outline for the practical, but some things had to be specified, like what

hormone(s) is/are used and how the concentration of hormones will be measured. In discussion with Jan-Henk Marsman from the analytic department of the RUG, the hormone choice became Beta- estradiol (also known as 17-β-oestradiol), simply because as seen in the LOES results (STOWA Derksen & Lahr, 2003) it is one of the hormones measured in surface waters and because it was in stock at the biochemical department of the RUG.

Jan-Henk also helped thinking of a way of measuring the concentration of beta-estradiol, which resulted in UV-spectrometrics. This is a fairly easy measurement method, which works because beta- estradiol has two absorption peaks, at 225 nm and 281 nm (Anon., sd).

3.2 Chemicals & Equipment

Chemicals: demi-water, ethanol (99%), beta-estradiol, coal (87% DSC)

Equipment: Heater, magnetic stirrer, measuring cylinder, beakers, spatulas, Buchner funnel

3.3 Method

Preparation

Put an amount of the coal in an oven overnight at approximately 100 °C to dry it. Weight before and after to control the DSC.

Method

First the PSD (particle size distribution) is determined of the coal. Pour an amount of the dried coal on a stack of sieves with each a different pore size (see image 2). The stack consisted of four sieves with pore sizes 0.025 mm, 0.050 mm, 0.100 mm and 0.200 m. Under the sieves a tray is placed to catch the smallest fraction (< 0.025 mm).But the stack in shake machine for ~30 min.

Now the particle size distribution of coal can be determined from the different fractions.

Image 5 Stack of sieves Image 6 Stirring setup

(24)

24 Next different hormone solutions are made. This is done in water and ethanol in the concentrations:

● 1 microgram/l

● 100 microgram/l

● 10 mg/l

● 100 mg/l

These concentrations are tested in the UV-spectrometric to see if they have a high enough

absorption to get clear results. The concentration with good results is taken as basis for the rest of the practical.

After choosing a suitable concentration, the following tests are run:

Test number Volume solution Amount of coal Type of coal Stirring time 1 200 ml 0.044 gram PS < 0.025 mm 1 hour 2 200 ml 0.063 gram PS <0.025 mm 24 hours 3 200 ml 0.402 gram PS < 0.025 mm 24 hours

4 200 ml 0.406 gram Mixture 24 hours

Table 4 Different tests for experimental

Every test had a blanco without hormone running with the same concentration of coal and same stirring time. This blanco is needed for the UV-spectrometric measurements.

After the stirring time every test had to be filtered thoroughly to make sure there is no coal left in the solution. This is done over a Büchner funnel. The same is done for the blanco solutions.

Image 7 Filter setup Image 8 UV-spectrometric

After this the UV-measurements can be done. Essentially two UV-measurements are done for every test.

1. This first measurement is done with a baseline of pure ethanol, and a measurement of the ethanol and hormone solution before adding the coal

2. The second measurement is with the filtered blanco as baseline and the filtered ethanol, coal and hormone solution as actual measurement

By comparing these two UV-measurements one can determine if hormone adsorption had taken place or not.

Two other smaller things have been looked at. The first one is the behavior of coal in water, this is simply done by putting 0,205 gram coal in 200 ml water. This is stirred for a few minutes and then the behavior is looked at.

Also the coal of test three has been looked at. By washing 0,203 gram coal that has been taken of the filter with 100 ml ethanol for several times (re-using the ethanol). This 100 ml ethanol will also be UV tested against a blanco of 100 ml ethanol. This is done to test the adsorption strength.

(25)

25

4. Results and discussion 4.1 Results

4.1.1 Overall

Although the literature stated that beta-estradiol solves reasonable well in water, there were some problems solving it in water. For the two lowest concentrations it worked after stirring for

approximately one hour while heating at 50 °C. These concentrations were not sufficient enough to get a nice peak at the UV-spectrometric. Because higher concentrations could not be achieved in water, the solvent was switched to ethanol. At 100 mg/l the absorption of light at the 225 nm peak was exceeding the 1.0 A, but for 281 nm it was around 0.75 A. This peak was used as reference for the delta measurement.

For the first three adsorption tests the fraction of the coal with the lowest particle size (< 0.025 nm) was used. For the fourth test the coal before sieving it was used, because that is the coal mixture that actually will be used at the WWTP. This means the concentration of the smallest fraction will be lower compared to the first three adsorption tests.

Image 9 Unsuccessful solving in water…

More tests could have been done, but the supply of hormone was not unlimited. This means choices had to be made.

-The first test was the amount of coal expected to be added according to Meelker´s report. He added the amount of coal to get a concentration of around the 0.2 gram coal per liter (Meelker, 2010). In this first test this concentration is achieved as well.

-Because the first test resulted in no adsorption noticeable, the amount of coal was slightly increased and the stirring time was significantly increased.

-The second test showed adsorption. This third test was to check if in the same amount of stirring time more adsorption can take place.

-The fourth test was about comparing the coal mixtures adsorption ability to the smallest fractions adsorption ability.

(26)

26 4.1.2 Coal

First the dry solid content (DSC) of the delivered coal was checked. The lid of the container of the coal stated a DSC of 87% and that it had to be determined again. For determining the DSC the assumption that after drying for 24 hours at 105 °C the DSC will be 100% has been made. This means that the weight difference before and after the drying is the weight of the water.

The DSC results are in table 4.

Mass before drying 6.65 gram

Mass After drying 5.87 gram

Difference(mass water) 0.79 gram

Dry solid content 88.2%

Table 5 Dry solid content

After drying and determining the DSC, the Particle Size Distribution was determined. See table 5, graph 3 and 4 and images 6,7,8,9 and 10 for the results. Particles larger than 0,100 mm make up for the majority (more than 60%) of the particles.

Particle size Mass (grams) Fraction of total Cumulative

< 0,025 mm 0.93 0.16 0.16

0,025 mm < PS < 0,050 mm 0.47 0.08 0.24

0,050 mm < PS < 0,100 mm 0.74 0.13 0.37

0,100 mm < PS < 0,200 mm 1.38 0.23 0.60

> 0,200 mm 2.35 0.40 1

Table 6 Particle Size Distribution

Graph 3 Particle Size Distribution 1 0

0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,91

> 0,200 mm

0,100 mm

< PS <

0,200 mm

0,050 mm

< PS <

0,100 mm

0,025 mm

< PS <

0,050 mm

< 0,025 mm

mass fraction

particle diameter

Particle Size Distribution

Fraction

(27)

27

Graph 4 Particle Size Distribution 2

Image 10 PS > 0.200 mm Image 11 0.100 mm < PS < 0.200 mm

Image 12 0.050 mm < PS < 0.100 mm Image 13 0.025 mm < PS < 0.050 mm 0

0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

0 0,1 0,2 0,3 0,4

Mass fraction

Particle diameter

Particle Size Distribution Coal

PSD

(28)

28

Image 14 PS < 0.025 mm

The behavior of coal in water gave the insight that most of the coal is settled after 40 minutes. Only a few small particles were floating around. These were probably the particles from the smallest

fraction.

4.1.3 β-estradiol

The results from the different tests (see experimental) are displayed and discussed in this section.

But first the graphs of the UV-measurements to find an optimum concentration to work with are shown.

The first three concentrations that were tried are:

● 1 µg β-estradiol/l water

● 100 µg β-estradiol/l water

● 10 mg β-estradiol/l water

The graphs of the UV-spectra are graphs 4,5 and 6.

Graph 5 UV-spectrum concentration test 1 µg/l 0

0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

200 250 300 350 400 450 500

Absorbance

nm

1 microgram /liter in water

1 microgram

(29)

29

Graph 6 UV-spectrum concentration test 100 µg/l

Graph 7 UV-spectrum concentration test 10 mg/l

As can be seen on the graphs, these concentrations were too low to see a clear light absorption peak.

This means the concentration had to be increased. As mentioned earlier, this was not accomplished in water. This resulted in trying the following two concentrations:

● 50 mg β-estradiol/l ethanol

● 100 mg β-estradiol/l ethanol

Graph 7 and 8 show the UV-spectra from these concentrations.

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

200 250 300 350 400 450 500

Absorbance

nm

0,1 mg per liter water

0,1 mg

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

200 250 300 350 400 450 500

Absorbance

nm

10 mg per liter in water

10 mg per liter

(30)

30

For the measurements of the tests the 100 mg β-estradiol / l ethanol concentration is used.

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

200 250 300 350 400 450 500

Absorbance

nm

50 mg per liter ethanol

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

200 250 300 350 400 450 500

Absorbance

nm

100 mg per liter ethanol

Blanco 100 mg per liter test 1

(31)

31 Test 1

Graph 10 Test 1 blanco

Graph 11 Test 1 result 0

0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

200 250 300 350 400 450 500

Absorbance

nm

Test 1: Blanco

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

200 250 300 350 400 450 500

Absorbance

Wavelength (nm)

Test 1: Result after 1 hour stirring

resultaat 100 mg per liter

(32)

32 Test 2

0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

200 250 300 350 400 450 500

Absorbance

Wavelength (nm)

Test 2: Blanco

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

200 250 300 350 400 450 500

Absorbance

Wavelength (nm)

Test 2: Result after 24h stirring

100 mg per liter ethanol result after 24h stirring

(33)

33 Test 3

0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

200 250 300 350 400 450 500

Absorbance

Wavelength (nm)

Test 3: Blanco

Test 4: Blanco

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

200 250 300 350 400 450 500

Absorbance

Wavelength (nm)

Test 3: Result after 24h stirring

Test 4: Result after 24h stirring

(34)

34 Test 4

Graph 17 Test 4 result

The important thing to take out these graphs is the light absorption value at 281 nm. The differences in these values between the blanco and the result can be calculated in to concentrations. These concentration differences resemble the adsorption of β-estradiol to the coal.

In table 7 the values of the peaks are shown.

Test number Peak value blanco Peak value result

1 0.7589 0.7562

2 0.7589 0.7150

3 0.7019 0.6451

4 0.9279 0.8860

Table 7 Peak values

These peak values can be calculated to concentrations via the Beer-Lambert law. This law relates the absorption of light to the properties of the material through which the light is traveling.

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

200 250 300 350 400 450 500

Absorbance

Wavelength (nm)

Test 4: Blanco

100 mg per liter in ethanol test 3 blanco

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

200 250 300 350 400 450 500

Absorbance

Wavelength (nm)

Test 4: Result after 24h stirring

Test 3: Result after 24h

(35)

35 To express this relation, the law states that a logarithmic dependence between transmission of light through a substance and the material, the distance the light travels through the material and the absorption coefficient of the material. The absorption coefficient is the product of the molar concentration and the extinction coefficient of the material.

For liquids this relation can be written as (Ingle & Crouch, 1988):

Where: T = Transmission, this is the fraction of the incoming light that comes through the material (-) I = The intensity of light measured after passing through the material (-)

I0 = The intensity of the incoming light (-) α = Absorption coefficient (1/cm)

l = Distance the light travels through the material (pathway) (cm) ε = Molar extinction coefficient (l /mol.cm)

c = molar concentration (mol/l)

This relation uses transmission; however the equipment used at this practical used absorbance (A).

The relation between absorbance and transmission is defined as (Ingle & Crouch, 1988):

With this relation the Beer-Lambert law can be written as:

Where: l = Distance the light travels through the material (pathway) (cm) ε = Molar extinction coefficient ( )

c = molar concentration ( ) A = Absorbance

The used cuvettes all had a pathway of 1 cm. The molar extinction coefficient ε has a value of 2040 . This means a concentration 100 mg per liter (must be in moles per liter, thus the concentration is divided by the molar mass: 272,38 gram per mole) will result in absorbance of:

Comparing this to the concentrations made at the tests, test 1 and 2 (which are from the same reservoir of solution) are quite accurate and around the 100 mg per liter. Test 3 and 4 are a little of set.

For every test the concentrations will be calculated from the absorbance, see table 8 for the data.

Test number Concentration blanco (g/l) Concentration result (g/l) Difference (g/l)

1 0.1013 0.1010 0.0003

2 0.1013 0.0955 0.0058

3 0.0937 0.0861 0.0076

4 0.1239 0.1183 0.0056

Table 8 calculated concentrations

(36)

36 Out of these concentrations the percentage adsorbed β-estradiol and the adsorption per gram coal can be calculated. See table 9 for the percentage adsorbed β-estradiol per test. These percentages are calculated by dividing the concentration differences from table 8 by the blanco concentration.

Test number adsorbed β-estradiol (%)

1 0.296

2 5.726

3 8.111

4 4.520

Table 9 Percentage adsorbed β-estradiol

To calculate the adsorption per gram coal, the concentration of coal is needed. Then the concentration difference (see table 8) is divided by the concentration of coal. In table 10 the concentrations coal and the adsorption per gram coal are shown.

Test number Volume solution

Amount of coal Concentration coal (g/l)

Adsorption per gram coal (g/g)

1 200 ml 0.044 gram 0.220 0.0014

2 200 ml 0.063 gram 0.315 0.0184

3 200 ml 0.402 gram 2.010 0.0038

4 200 ml 0.406 gram 2.030 0.0028

Table 10 Adsorption per gram coal

Also the strength of the adsorption was tested. This resulted in the UV-spectrum in graph 17.

Graph 18 Adsorption strength test 0

0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

200 250 300 350 400 450 500

Absorbance

Wavelength (nm)

Adsorption strength test

Bachelor thesis Chemical Engineering